The WS Contest tool was breaking on Index pages that don’t yet have any (existing) pages. The problem was in the usage of the dflydev/dot-access-data, which is a thing for pulling values out of deep arrays. It’s sort of useful, but not widely used so does feel a bit weird. Maybe we should remove it from wikisource/api.
Actually, scratch that: I just went to look, and it’s got eighty-three million installs on Packagist. So I guess it’s more widely used than I’d thought. I’ll leave it in for now.
I wonder if the wool wagon (at the left background) in this photo of a cricket match at Wooramel in about 1930:
is the same one seen here in 2020:
Hacker News is going on about the CSS of the W3C’s homepage not working. I was worried, reading the headline, that they’d got rid of the Core Styles! But no, those are safe.
I’ve been working on linking Commons items to Trove. It’s not complicated, on the surface, but of course once you get into it things get more annoying. I’m trying to just figure out the basics, and not get too deep into the weeds.
Basically what we want is for every file on Commons, that has been sourced from an institution that’s aggregated in Trove, to have a P10044 statement (‘Trove Work ID’), and for that to be displayed in the |source= parameter of the Information template. But we also want to say which library (etc.) it came from, and to link to that library’s catalogue (because that’s where the actual image file is stored). This means, I think, that we’ll want to stick to the SLQ, SLWA, SLSA, AWM-image, etc. templates that have institution-specific information and links, and just add a simple line to those that explains that the file’s metadata is also on trove and what its ID there is.
The Community Wishlist Survey starts today. This is an annual survey that my team at the Wikimedia Foundation runs, to figure out what we should work on for the following year. It’s always so interesting to see what features and bugs are felt to be the most important. Quite often I agree, and get excited about working on them; sometimes I’ve never even heard of the software components in question! The Wikimedia universe is wide and varied.
So if anyone Wikimedians out there have ideas, please come and propose them!
The (fairly terrific) locator-tool now has aerial imagery available as a map layer, making geolocating photos even more fun.
A new extension this week: CopyCreds
The CopyCreds extension registers two new tags <pw> and <un>, which make the text inside them visually distinctive, and allows for click-to-copy. The goal is to make life easier for those who document usernames and passwords in MediaWiki.
I like the idea of a simple click-to-copy system for a wiki, but I’m not sure it warrants two new tags. Personally I’d prefer something like <copy class='pwd'>foo</copy> or something like that, with a means to customize the colour etc. of the displayed value. Actually, maybe that’s what CopyLink does. There used to be one called CopyToClipboard which also did something similar.
Maybe someone should write the Copyencabulator 2000, to unite them all…
I’ve seen a few people talking about Thunderbird recently, and the idea of email in general. I like email, for private communication — meaning pretty much only one-to-one. For public or group stuff it’s pretty dreadful.
Desktop email, and from Alpine to Thunderbird by Ruben Schade, 20 January 2023:
Shocking nobody who’s dismissed me as a contrarian before, I still use desktop email clients. Even worse, I only made the switch to IMAP from POP for personal email a few years ago, in part owing to the principle of if it ain’t broke, don’t fix it. […] They let me archive email locally, which reduces my footprint on a remote third party server somewhere. It’s better for privacy, and potentially cost. […] HTML email though, that’s something I still refuse to budge on… because I’m a gentleman.
The matter of how email is stored, where, and for how long is interesting I think. I’ve been reading (and in some cases scanning and putting online) some old letters from long-dead relatives of mine. It’s an amazing way to hear someone’s voice, and I’m glad that various people in my family have had stable homes (for many decades at a time) and the inclination to not throw things away. It makes me wonder (and I know this is a cliché) how emails and other personal communication will survive the coming decades. It’s just so much easier to lose access, not care, or not even notice when things get deleted.
I used to maintain an email archiving and sending tool for myself. It was plain text, didn’t allow attachments, and produced annual PDF dumps of conversations (unthreaded, purely chronological, and only between two parties). Utterly limited, but I found it not only terrific to use, but it encouraged me to write longer and more detailed emails. The explosion of chat systems came along though, and most of my friends moved to other platforms and didn’t want to email any more, so I’ve let the thing stagnate. Maybe I’ll make it live again some day.
In reply to this post by Sam Wilson.
It does exist, as of just now.
I can’t believe there’s no ‘Ghost signs in Western Australia’ category on Commons!